In three separate studies involving three different sets of collaborators, elevated levels of 2,3-butanediol have been found in the blood of 80% of chronic alcoholics, but not social drinkers consuming distilled spirits. Two separate methods of gas chromatographic analysis of diols have been developed. One method involving formation of the bromophenylboronate derivative can accurately measure to d-1, or meso 2,3-butanediol to 5 uM. In the rat, two pathways of butanediol formation have been demonstrated. The first (Veech et al.: Curr Top Cell Regul 18: 151-179, 1981) involves elevated blood acetaldehyde entering the brain with an active pyruvate dehydrogenase multi-enzyme complex where it condenses with hydroxyethyl thiamine pyrophosphate to form acetoin. The acetoin is subsequently converted in liver to 2,3-butanediol. In a second animal model, 2,3-butanediol in the rat is produced by acetone feeding. Prolonged fasting in man, however, produces diols but not 2,3- butanediol, suggesting differences in the metabolic pathways between man and the rat. Whether 2,3-butanediol in blood may be used as a "genetic marker" for alcoholism or is an induced defect can only be determined by a controlled study involving alcohol administration under controlled conditions on a metabolic ward.